Leisure, Consumption and Long Run Risk: An Empirical Evaluation

Abstract

I study a long-run risk model with non-separable leisure and consumption in the Epstein-Zin preferences to price a cross-section of equity returns over 1948-2011 for the US market. The stochastic discount factor is shown by news on both leisure and consumption. While estimating these two long run factors using a vector auto-regression (VAR), I find that growth (big) stocks - lower average returns - obtain higher long run leisure betas but lower long run consumption betas than value (small) stocks do. Here leisure acts as an 'insurance' and decreases the price of long run risk. Expected returns on stocks that highly comoves with future leisure will be less because of future investment opportunities. Moreover, expected cash flows are small if there is good news on leisure. The data fact evidences that leisure absorbs more than 60.2% foregone work hours, but less than 2% for searching jobs. When job creation flows hampers for the marginal costs of hiring fails to decline fast, the rate at a vacancy is filled decreases, therefore expected cash flows become even smaller as productivity falls and wages are inelastic.